恙蟲病立克次體 Orientia tsutsugamushi 菌株 TW-1 和 TW-22 之全基因組: 以潛在毒力因子與標的基因來進行菌株的鑑定和定性

Abstract

背景：恙蟲病的病原體是恙蟲立克次體 (Orientia tsutsugamushi, OT)，是一類專性細胞內共生的細菌，主要透過恙蟎來傳播。臺灣全島皆有恙蟲病個案，嚴重時會致死，由於每年確認的病例數高，長年來被視為最普遍、最嚴重的蟲媒傳染病之一。現今 OT 的菌株分型主要根據表面抗原 56-kDa 特異性抗原 (TSA56) 的基因序列多型性，近年來臺灣臨床上的 OT 分離菌株序列亦呈現其多樣性。由於缺乏其全基因組序列的資訊，目前亦不清楚 TSA56 的變異是否能反應整個基因組序列的差異，包括其他表面抗原和毒力因子的差異。此外，臺灣恙蟲病病例的臨床表現呈現極大差異，是否與 OT 菌株差異所表現之毒力有關則仍不清楚。 目標：本文研究臺灣最常見的 OT 菌株 TW-1和 TW-22，期能 (1) 構築兩者之全基因組序列後，分別進行潛在毒力因子的比較性研究；(2) 鑑別代表性的抗原蛋白標的基因，使其具備與OT菌株基因組層級親緣關係有相同解析力； (3) 透過研究具感染性的OT 菌株探討臺灣恙蟲病的早期臨床表現。 方法： 將 OT 菌株 TW-1 和 TW-22 培養於小鼠 L929 細胞中。建構OT 基因組之前，首先過濾純化菌體並去除小鼠細胞 DNA，然後以 PacBio Sequel 高精準長讀取定序系統進行全基因組定序，再使用 hifiasm、circlator、pilon 等常用的開源軟體進行 de novo 組裝，最後使用 Illumina MiSeq 平台讀取進行校正並拼裝。以 NCBI 原核基因組註釋流程(PGAP)進行基因組註解，並使用 BLAST+ 演算工具和 SMART web 伺服器比較 TW-1 和 TW-22 基因組中潛在的毒力因子。鑑定 OT 全基因組的核心氨基酸序列後，以鄰近連接法 (NJ)、最大簡約法 (MP) 和最大似然法 (ML) 分析基於基因組核心和個體抗原蛋白序列建立的親緣關係樹。另一項研究則使用邏輯回歸分析了與臺灣疾病管制署 10 年間 OT 分離株相關之早期臨床表現，以及各地區感染 OT 菌株類型的關係。 結果：使用覆蓋率超過 100倍的 PacBio HiFi 短序列 (reads)、組裝並獲得 OT TW-1 和TW-22 菌株的環形全基因組序列，並且使用 Illumina MiSeq 平台讀取檢視下沒有被進行糾正。TW-1的長度為 2,008,429 bp，包含 2067 個註釋基因，其中包括 1627 個完整的編碼序列和 400 個偽基因，而 TW-22 的長度為 2,044,475 bp，含有 2192 個註釋基因，其中包括 1738 個完整的編碼序列和 414 個偽基因。相較於 TW-1，TW-22 含有更多重復的蛋白質，包括含有 ankyrin repeat (Ank) 結構和tetratricopeptide repeat (TPR) 結構的蛋白質。在親緣關係上，對 691 個核心氨基酸序列進行分析，在 NJ 和 ML 模型中得到相同的拓撲結構。在 TSA56 和 Sca 蛋白 (ScaA、ScaC、ScaD 和 ScaE) 中，僅使用 TSA56 進行的 NJ 和 ML 樹的拓撲結構與核心基因組的親緣關係樹形相比大致相似，但仍然存在不一致性。然而將 TSA56 和 ScaA 的氨基酸序列連結後的親緣關係樹形，則與核心基因組的親緣關係樹形有高度相似的拓墣結構，而 TSA56 和 ScaD、ScaE 序列連結使用的拓墣結構也比單獨使用 TSA56 時更似於核心基因組的親緣關係樹形。臺灣南部地區，臨床上相較於其他菌株，TW-22 的患者出現焦痂的幾率更高。 重要性：本研究首次提供臺灣 OT 的全基因組序列，有助於解釋全球 OT 菌株的多樣性。一些證據呈現恙蟲病臨床表現與 OT 毒力基因或菌株層級有相關聯，如體外培養實驗之生長動力學和潛在的分泌效應基因上的數量差異，因此需要進一步研究是否與菌株的毒力有關。與全基因組的親緣關係相比，特性化 OT 菌株並使用串聯 TSA56 和 ScaA 的序列分析，在親緣關係建立上得到顯著的突破，這可能對 OT 分離株的特性具有臨床價值。儘管在臺灣恙蟲病早期臨床表現的差異上，從分離株層級中找到有限的證據，但與疾病嚴重程度相關的焦痂在分離株層級上仍需要進一步研究。

Background: Orientia tsutsugamushi (OT) is an obligate intracellular bacterium associated with trombiculid mites. OT is the causative agent of scrub typhus, a life-threatening disease with an expanding range of endemicity and the most prevalent vector-borne disease endemic to Taiwan. 56-kDa type-specific antigen (TSA56) sequence-based strain typing of OT has revealed remarkable strain diversity, especially among clinical isolates in Taiwan. However, genome-wide variation among OT strains remains poorly characterized, including for diverse sets of secreted protein effectors that represent putative virulence factors and other antigenic proteins that have yet to be evaluated for their ability to reflect core genome-based phylogeny. Furthermore, it remains unclear whether OT strains vary in clinical presentation and disease severity. Aims: This dissertation aimed to (1) obtain complete genomes for two common OT strains in Taiwan, TW-1 and TW-22, for comparative study of their predicted secretomes, (2) identify antigenic protein sequences that reflect core genome-based phylogeny, and (3) investigate early clinical manifestations of scrub typhus in Taiwan by infecting OT strain type. Methods: OT strains TW-1 and TW-22 were cultivated in mouse L929 cells, and DNA extracted from purified OT was used for whole genome sequencing using PacBio Sequel for de novo assembly and Illumina MiSeq for polishing, performed using widely-accepted open-source software, including hifiasm, circlator, and pilon. Assemblies were annotated using the NCBI Prokaryotic Genome Annotation Pipeline, and amino acid sequences in TW-1 and TW-22 genomes were compared using BLAST+ and the SMART web server. Core amino acid sequences were identified for complete OT genomes. Neighbor-joining (NJ), maximum parsimony (MP), and maximum likelihood (ML) models were used to examine core genome-based phylogeny and phylogeny of individual antigenic proteins. A retrospective study examined early clinical manifestations associated with OT isolates at the Taiwan Centers for Disease Control over a 10-year period in relation to infecting OT strain type in each region using logistic regression. Results: Complete circular genome sequences were obtained OT strains TW-1 and TW-22 with >100 coverage using PacBio HiFi reads and no corrections with Illumina MiSeq reads. TW-1 was 2,008,429 bp and 2067 annotated genes, including 1627 intact CDSs and 400 pseudogenes, and TW-22 was 2,044,475 bp with 2192 annotated genes, including 1738 intact CDSs and 414 pseudogenes. TW-22 had more duplicated ankyrin repeat-containing (Ank) proteins and tetratricopeptide repeat-containing (TPR) proteins than TW-1. Phylogenetically, analysis of 691 core amino acid sequences produced identical topologies for NJ and ML models. Among TSA56 and Sca proteins (ScaA, ScaC, ScaD, and ScaE), NJ and ML tree topologies using TSA56 were comparatively most similar to the core genome phylogeny, though incongruent. Use of concatenated TSA56 and ScaA amino acid sequences resulted in highly similar tree topologies with the core genome phylogeny, while TSA56 and ScaD or ScaE also produced more similar tree topologies than TSA56 alone. Clinically, in southern Taiwan, TW-22 was associated with higher odds of presenting with an eschar compared to other strains. Significance: This work contributes the first complete genome sequences of OT in Taiwan, addressing a critical geographic gap. Strain-level differences were observed in the number of intact duplicated secreted effectors, necessitating further studies to determine correlation with strain virulence. Compared to genome-wide phylogeny, OT strain characterization was markedly improved using concatenated TSA56 and ScaA sequences and may be useful to characterize OT isolates. Limited evidence was found for strain-level differences of OT in early clinical manifestations of scrub typhus in Taiwan, but strain-level differences in eschar development should be further studied in relation to disease severity.